This study was undertaken in an attempt to determine the differential effects of single and double response channel sets learned to fluency (rates of 50–60 correct associations per minute) versus overlearning (90% correct or greater with 3-second latency between associative responses) on subjects' retention of paired associate information. An intrasubject multiple treatment design was used, exposing each subject to all training conditions: baseline and mastery-learning sequentially, followed by the repeated “multiple treatment” exposure to four conditions; 1 and 2 channel fluency, and 1 and 2 channel overlearning conditions. Retention was measured immediately after training, and at one, two, three, seven, and twelve week follow-up intervals. Additionally, the duration of time and number of learning trials required for subjects to meet mastery, fluency and overlearning criteria at both 1 and 2 channels was calculated to gather further information regarding whether or not the magnitude of differences is educationally significant. Moreover, subject satisfaction was measured in response to the mastery, fluency and overlearning conditions. The results indicated no significant differences in retention produced by any of the 4 instructional conditions, at any follow-up interval. Additionally, the results indicated that fluency goals require an “educationally significantly” greater number of learning trials and more time to attain than mastery-based accuracy goals. What's more, it appears that adding a second response channel during fluency building does not seem to accelerate fluency development, as compared to a 1-channel fluency condition. Furthermore, subjects appeared to prefer faster fluency based training over slower overlearning training. It is suggested that future research utilize longer term follow-up intervals and higher fluency rates.